Offspring sex ratio is affected by pre-breeding rainfall and hatching order in a cooperative breeding bird

Selection may favour sex ratio adjustment when the fitness benefits of producing male or female offspring vary. In birds, sex ratio bias at hatching has been particularly well-studied in cooperative breeding species, primarily in helper-at-the-nest systems wherein young remain with their parents and help raise subsequent broods. In these systems, breeding females are expected to skew offspring sex ratio in favour of the helping sex if sufficient resources exist. In communal laying systems multiple group members breed and raise the young and offspring dispersal varies, making it difficult to predict whether and how breeders should skew offspring sex ratio. We tested for sex ratio bias in smooth-billed anis (Crotophaga ani), a communal laying cuckoo with low within-group relatedness, high offspring dispersal, and high annual turnover in group membership. One male performs nocturnal incubation and sires more offspring, suggesting these males may have higher reproductive variance and thus that it may be beneficial for his mate to produce more male offspring. We hypothesized that pre-breeding rainfall influences food availability and thus offspring sex ratio, predicting that breeding females skew production towards the more beneficial sex (i.e., males) in high food years, consistent with the Trivers-Willard hypothesis. Females may also adjust sex ratio across the hatching order, so we predicted a male-bias in first hatched chicks in larger broods that should experience greater competition. Contrary to our first prediction, more male chicks hatched when pre-breeding rainfall was lower. In partial support of our second prediction, marginally more first hatched chicks were male in larger broods. Based on these results, we propose the nestling competition hypothesis and predict that when nestling competition is high (i.e., in poor years and in large joint broods), mothers should produce more of the more competitive sex.

Methods
Data were collected from wild breeding smooth-billed anis at a single study site (Cabo Rojo National Wildlife Refuge) over 5 breeding seasons (2005, 2011-2014). Nests were located and monitored regularly throughout each breeding season to determine the beginning of egg laying, onset of incubation, beginning of hatching, and end of hatching. We used molecular techniques to determine the sex of each chick.We used rainfall data collected from a U.S. Fish and Wildlife Service weather station on the Cabo Rojo NWR (17°58’23" N, 67°09’46" W; data available at https://wrcc.dri.edu/cgi-bin/rawMAIN.pl?txPCAB). Rainfall data from this station during the breeding season were not available for 35 study days, largely due to the U.S. government shutdown in 2013. To determine rainfall for these days, we used data from the USGS-NRCS/SCAN weather station near the main entrance to the refuge (17°59’00’’N, 67°10’00’’; data available at https://wcc.sc.egov.usda.gov/nwcc/site?sitenum=2066). Data organization and analyses were conducted in R version 4.0.3 (R Development Core Team, 2021), using base R and the packages dplyr (Wickham et al., 2023) and lme4 (Bates et al., 2015). Data visualization was done using the package ggplot2 (Wickham, 2016). We followed all applicable international, national, and institutional guidelines for the care and use of animals in research.